Releases: Ricks-Lab/benchMT
Releases · Ricks-Lab/benchMT
v2.0.0 Major Release
- Complete rewrite of the core to integrate GpuItem object that was previously only used in energy mode.
- Implemented a more robust check of compute and energy compatibility for all installed GPUs. Switched from lshw to lspci and clinfo. Not working for Intel GPUs.
- Many core improvements for better robustness and style. Added reST docstrings for better supportability and improved type checking by PyCharm.
- Extend energy mode to cover Nvidia, in addition to AMD GPU's. Added max-power to psv and txt reports.
- --noBS on the command line will allow benchMT to function without a BOINC installation.
- Check devmap and give more details to help user understand mapping of BOINC Devices to Driver Card Numbers.
- Added --purge_kernels command line option to purge cached kernel files from working directory.
- Added --lsgpu command line option to display information on all installed GPUs.
- Reading BOINC's coproc_info.xml as a possible way to automate device to card mapping. Not implemented.
v1.6.0 Feature Release
- Complete rewrite of commandline/config file option parsing. Original got complex and buggy.
- Support execution and time/energy metrics for AstroPulse apps/wus. Still no working results comparison utility, so comparison to reference results not possible.
v1.5.0 Feature Release
- Implemented checks of Python and Kernel version to verify compatibility.
- Implemented more robust system calls.
- Fixed an error in parsing lshw output to extract GPU names and added error checking.
- When specifiying specific GPUs to be used with --gpu_devices, max_gpus is now automatically set.
- Implemented Energy metric for GPUs using amdgpu drivers.
- Code robustness improvements (path joins and key checks).
- Run completion message now includes the location of the results.
v1.4.0 Feature Release
- Write run_name to the psv file. Useful when wanting to analyze the data from multiple runs.
- Include nVidia stock MB app in the distribution.
- Include error message when job fails to spawn.
v1.3.0 Feature Release
- SETI MB apps are now run without a shell, using shlex to parse args for the subprocess command.
- Implemented --force_ref option to force generation of reference results, even if they already exist.
- Implemented job execution with time command. Time relevant data is written to summary and psv files.
- Added job execution error checking. Bad exit status will result in updated error fields in summary/psv files and status display.
v1.2.0 Features and Bug Fixes
- Fixed a problem with the when lock_file was created and checked. Now placed before slot initialization.
- Fixed issue where program would exit if Reference file didn't exist. Now an error message is printed and no comparison results are printed to summary files.
- Added commmand line option --no_ref which will not create reference results when selected. This is useful for characterizing potential reference WUs.
- Added color to status display.
- Modified so that status display will not show skipped jobs (Reference data already exists).
- Updated reference WUs in the WU_test/safe directory. Still need a WU with a Gaussian signal.
Configuration Features Release
- Command line options can now be specified in mode lines of the BenchCFG file. Options given on the command line will override modes specified in the CFG file.
- An alternative CFG file can now be specified as a command line option.
- Signal counts and Angle Range are now included in the psv and txt summary files.
- Remove app -device N arg if specified, since -device is automatically added based on slot assignment.
- Added --gpu_devices x,y command line option to specify which GPU devices the user would like to include in the benchmark run.
- Added a lock_file in the working directory to prevent a second occurrence of benchMT from using the same directory.
- Updated the 15 reference WUs in the WU_test/safe directory.
- Changed --ref_signals option to --std_signals for clarity.
Initial release
This is the initial release for the project.